This commentary piece is an adapted version of my discussion from The Week That Was on January 31, 2025.
#TWTW is a weekly X Spaces show hosted by D.R.Lewis and Matt Case, airing every Friday at 4 PM EST. It covers the hottest news about Bitcoin mining stocks —delivered in a fun and engaging way. Tune in if you can!
A little-known Chinese AI startup, DeepSeek, has suddenly become the center of attention in the tech and investment world. Within a week, it ranked as the No. 1 app in App stores across Australia, Canada, China, Singapore, the U.S., and the U.K. At the same time, nearly $600 billion was wiped off Nvidia’s market value overnight, sending shockwaves across the AI and Bitcoin mining sectors.
All publicly traded Bitcoin mining stocks tracked on our platform also took a hit, reflecting the uncertainty surrounding the future of AI infrastructure spending. The buzz around DeepSeek is intense because everything happened so fast, and people have a lot of questions. So, let’s break it down and go over what’s happening.
Who is DeepSeek AI?
DeepSeek AI is a Chinese artificial intelligence company based in Hangzhou. Its founder, Liang Wenfeng, also funded one of China’s top hedge funds, High-Flyer, which focuses on AI-driven quantitative trading.
In 2023, DeepSeek AI was founded because High-Flyer wanted to put its idle GPU capacity to use. The company released its first model in November 2023 and has since iterated multiple times on its core large language model (LLM), building several variations. However, it wasn’t until January 2025—after releasing its R1 reasoning model—that DeepSeek AI gained global attention.
This model is claimed to rival OpenAI’s GPT-4o in performance benchmarks. Its key feature? Advanced reasoning skills, such as the ability to rethink its approach to math problems.
On top of that, it’s significantly cheaper than OpenAI’s GPT-4o. How cheap? A quick comparison:
- DeepSeek trained R1 using only 2,000 Nvidia H800 chips and spent just $5.5 million.
- OpenAI’s GPT-4o cost over $100 million to train and needed some of the most powerful Nvidia GPUs available.
How DeepSeek AI Disrupts High-Performance Computing and Bitcoin Mining
DeepSeek AI’s rollout is shaking up the market. Investors are now questioning whether U.S. companies’ massive AI spending is justified. If a company like DeepSeek AI can build highly competitive AI models at a fraction of the cost, does it mean we’ve been overestimating the amount of computing power needed?
This has dramatically changed market expectations about the demand for computing power. The conventional belief was that developing cutting-edge AI models required huge amounts of high-end hardware. But DeepSeek AI is proving that more can be done with less.
As a result, investors immediately priced in the disruption that a low-cost AI competitor could bring. Stock prices of chipmakers and tech giants—Microsoft, Alphabet, Amazon—plummeted. These companies have spent billions developing AI tools like ChatGPT and Gemini, so any sign of a shift in AI economics raises big concerns.
And since tech stocks make up about 45% of the S&P 500, the broader market also took a hit. The initial reaction to uncertainty is always to sell.
But Here’s the Thing...
We don’t actually know how advanced DeepSeek AI’s model really is.
According to White House AI and crypto czar David Sacks, there’s “substantial evidence” suggesting that DeepSeek AI may have distilled knowledge out of OpenAI’s models. He added, "I don’t think OpenAI is very happy about this."
And there’s more skepticism. Bernstein analyst Stacy Rasgon dismissed DeepSeek AI’s claim that it only spent $5.6 million on training costs. He called it "categorically false," pointing out that it doesn’t account for other substantial costs involved in AI model development.
So, if the claims about DeepSeek AI’s efficiency turn out to be exaggerated, the panic selling we saw this week could be an overreaction.
Long-Term Outlook for AI Infrastructure and Bitcoin Miners
If AI can truly be developed at a lower cost, that could accelerate innovation—not slow it down. More companies would be able to enter the space, expanding the market faster than expected. In the long run, the value of AI infrastructure will continue to grow.
Let’s not forget: President Donald Trump recently announced a $500 billion investment in AI infrastructure through a joint venture known as Stargate. That’s a massive commitment to long-term AI development. On January 30, Cipher Mining announces $50 million PIPE investment from SoftBank. All these indicates large tech players are coming and coming faster.
So, while the market reacted strongly last week, I see this as an overreaction. The long-term demand for high-performance computing (HPC) isn't going away and will keep growing—and that’s good news for Bitcoin miners with HPC/AI exposure.
Disclaimer: The views expressed in this article are my own and are based on publicly available information. This content is intended for informational purposes only and should not be construed as investment advice. Readers are encouraged to conduct their own research before making any investment decisions. Past performance is not indicative of future results. No recommendation or advice is being provided as to the suitability of any investment for any particular investor.
